BACHELOR THESIS: TESTING AND IMPLEMENTATION OF VARIOUS SELF DRIVING CAR FEATURES USING NVIDIA JETSON NANO AND JETRACER PRO AI KIT
The main goal of our project is to implement self driving car features namely lane detection and following and obstacle avoidance using Python and its Machine learning and deep learning libraries as it is has a more simple and consistent environment with advanced AI/ML framework support. Our secondary objective would be to achieve ob stacle avoidance using SLAM (Simultaneous Localization and Mapping) and ROS (Robot Operating System) with the help of a prototype which is a 1:18 model of standard car. We made use of NVIDIA JETSON NANO which is a high performance board enabled with deep learning and Artificial intelligence capabilities. We have implemented our first project objective, lane following and obstacle avoidance using Machine Learning and Deep Learning through Python and its libraries. The next goal of the project was to implement obstacle avoidance using ROS (Robot Operating System), for that we have implemented SLAM using Google Cartographer after a thorough comparison between a lot of SLAM implementation methods. For more information about the project please refer to our detailed project report Major_project_report_final.pdf